wav2vec2-lv-60-espeak-cv-ft-WCTCv2-phocab-ds-f0

This model is a fine-tuned version of facebook/wav2vec2-lv-60-espeak-cv-ft on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 106.1334
• Per: 0.2026

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Per
1374.3028	0.7194	400	391.5120	1.0
521.4664	1.4388	800	379.8089	1.0
497.757	2.1583	1200	351.1022	1.0
376.6007	2.8777	1600	163.3355	0.4681
204.1381	3.5971	2000	125.0853	0.2898
159.1483	4.3165	2400	121.2204	0.2716
137.1822	5.0360	2800	117.1236	0.2557
120.3087	5.7554	3200	115.2592	0.2489
112.0666	6.4748	3600	116.7352	0.2451
107.1429	7.1942	4000	111.5922	0.2405
105.1262	7.9137	4400	114.3477	0.2382
98.853	8.6331	4800	110.9513	0.2307
94.6806	9.3525	5200	114.5719	0.2291
94.9273	10.0719	5600	112.2022	0.2246
90.8426	10.7914	6000	108.0653	0.2208
86.705	11.5108	6400	108.8957	0.2284
90.1368	12.2302	6800	106.8571	0.2200
85.8211	12.9496	7200	107.4370	0.2140
86.2175	13.6691	7600	106.5738	0.2109
86.3826	14.3885	8000	111.6841	0.2140
82.317	15.1079	8400	110.4203	0.2155
82.7148	15.8273	8800	109.2693	0.2162
82.6152	16.5468	9200	103.8936	0.2140
81.6078	17.2662	9600	105.5971	0.2071
79.4881	17.9856	10000	105.5673	0.2086
79.041	18.7050	10400	106.5618	0.2018
80.3571	19.4245	10800	105.3801	0.2094
78.0549	20.1439	11200	109.1485	0.2132
77.44	20.8633	11600	105.7328	0.2041
74.5847	21.5827	12000	103.5090	0.2049
76.4431	22.3022	12400	105.0128	0.2071
74.5507	23.0216	12800	106.0122	0.2102
74.0633	23.7410	13200	106.0005	0.2018
75.3061	24.4604	13600	106.4651	0.2094
73.7444	25.1799	14000	106.6356	0.2056
73.9277	25.8993	14400	105.9167	0.2056
73.1383	26.6187	14800	105.9251	0.2049
73.1463	27.3381	15200	106.9326	0.2056
72.6878	28.0576	15600	106.4586	0.2064
71.3269	28.7770	16000	106.0886	0.2064
72.607	29.4964	16400	106.1334	0.2026

Framework versions

• Transformers 4.57.6
• Pytorch 2.9.1+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2